Print Page | Close Window

Power Amp Repair - fuses

Printed From: Speakerplans.com
Category: General
Forum Name: Amp Forum
Forum Description: The 'Stopping Jake Fielder moaning constantly' forum description...
URL: https://forum.speakerplans.com/forum_posts.asp?TID=108250
Printed Date: 20 April 2024 at 10:54am
Software Version: Web Wiz Forums 12.06 - https://www.webwizforums.com


Topic: Power Amp Repair - fuses
Posted By: spacebiscuit
Subject: Power Amp Repair - fuses
Date Posted: 28 December 2023 at 11:16pm
I have a Mosfet T-400 Power Amp developed an issue with the volume control so I switched the Potometer. during the change I damaged the tracks a little so soldered in some wires to address this. After reapir I tested on the bench and was ready to reassemble the channel into the main unit.

I then did something really stupid and reversed the polarity, when i switched the amp on there was a brief moment of smoke before I switched the power off but clearly damage has already been done. The amp has two replaceable fuses which have a red and green led. When I power the amp up the green fuse blow immediately and there is then a repeated pulsing sound coming from the speakers.


I've used a multimeter to check the diode (seems there is only one), the transistors and resistors and caps. There is no noticeable sign of burnt out components and the readings I am getting tally with the corresponding component which I am verifying against on the good/working channel.

the smoke looked like it was coming from the bottom right hand corner where the small additional board which is labelled as "cuttoff" sits (detached in photo with brown, orange and green wires). I was convinced this was the issue but switched it over to the good channel and it worked so seemingly this is not the issue.


The small board I swapped out:



This is the rear of the board I repaired after I damaged the tracks:



I'm kind of out of ideas, I have used my bag of spare fuses so my trial and error approach to repair is not working.

I have not removed any components from the board to test with the multi-meter - should this matter?

Thanks in advance.




Replies:
Posted By: kedwardsleisure
Date Posted: 28 December 2023 at 11:47pm
It's not the most complicated board, why not remove and test all the semi's, store them safely then go over the resistors with the multimeter in situ, lifting any that dont test.

Do you have the facility to test the mosfets?

-------------
Kevin

North Staffordshire



Posted By: spacebiscuit
Date Posted: 29 December 2023 at 4:26pm
I've removed two of the MOSFET, I tested with a multimeter:

Red probe to Source (center pin), black probe to D (multimeter beeps) then to G and multimeter beeps again. If I am not mistaken this indicates a damage?

The piece is "J351" i guess this:

https://www.web-bcs.com/transistor/tmfet/j0/J351.php



Posted By: EmPro
Date Posted: 02 January 2024 at 8:03am
mosfets are done,buy new ones,don't bother with those ones anymore.

-------------
EmPro Audio


Posted By: kedwardsleisure
Date Posted: 03 January 2024 at 3:48pm
difficult to test mosfets properly with a multimeter, really you need a component analyser or tester, this is because on a good one the gate can float and turn the d-s junction on, making them look like they're short when they're not. If in doubt, just replace them


-------------
Kevin

North Staffordshire



Posted By: spacebiscuit
Date Posted: 18 January 2024 at 6:43pm
I ordered some replacements P-MOSFET's - I fitted them but unfortunately now both fused are tripping when I power up the amp - before the repair just the fuse with green led was tripping.

Not sure if the MOSFETs were poor quality or faulty - they came from China as I can't find them stocked in the UK.

With MOSFETs removed the fuses do not trips when I power up the amp.

I've ordered more of the P-MOSFETs and a cheap component tester (LCR-TC7) to test the MOSFET before I try to fit them.

I also have ordered some N-MOSFETs (board has 2 of these) but I'm not sure if the current pair need replacing and I'd like to avoid removing them un-necessarily if they're ok but I'm not how else I can test them.

Any thoughts or ideas?




Posted By: EmPro
Date Posted: 18 January 2024 at 8:51pm
Mosfets from china...fake as hell...i bought just once some V-FET's that supposed to be original from a dead power supply amplifier...but i received some odd like new mosfets...they blew up on start up right away,never going to do that mistake again,but in your case just wait for the tester to arrive and try to find genuine parts...

-------------
EmPro Audio


Posted By: kedwardsleisure
Date Posted: 18 January 2024 at 8:59pm
the mosfets are almost certainly going to be fakes, what sort are they, renesas? Are they laterals or v-fets? Laterals are often faked and v-fets you should be able to get from a franchised distributor.

-------------
Kevin

North Staffordshire



Posted By: spacebiscuit
Date Posted: 18 January 2024 at 10:12pm
Absolutely no doubt that they are fake, but reassuring to know that others have had a bad experience with these because it hopefully indicates that there is not another problem with my amp other than crappy replace. Ent parts.

Not sure of the type but datasheet:

https://www.alldatasheet.com/view.jsp?Searchword=2sj352&gad_source=1&gclid=Cj0KCQiAtaOtBhCwARIsAN_x-3KQYfu6nbf2rsHy5yRkj0QEEbWaDo-CfMM8i3hH9k96EWfls6wi_5gaAmXLEALw_wcB

I have looked but can't find any of these from UK sellers, seems they are discontinued hence my only option to data has been the Chinese import. 




Posted By: M4trix
Date Posted: 19 January 2024 at 12:21am
You might try Exicon's transistors such as ECX10N20/ECX10P20. Similar specs, a bit faster than 2SJ352/2SK2221. Either way, you don't have many options since they're way obsolete. Renesas still has the option to buy them but only if you order couple of thousands. LOL

Here: https://gb.profusion.uk/uk/ecx10n20" rel="nofollow - https://gb.profusion.uk/uk/ecx10n20

Edit: Check the other components thoroughly or you'll have an expensive damage again. Ouch
Btw, what's with the mounting hole on the right side of the board ?


Posted By: kedwardsleisure
Date Posted: 19 January 2024 at 10:40am
yes those are renesas lateral fets, highly faked.

I remember my first fake one of these, well before buying from china/ebay was a thing, I only needed one and it came from a tv component supplier (remember them?). It went into a C-Audio amp, I forget the model, and when I switched it on the legs blew off (the amp kept working).

I think my last batch of these were supplied by Matrix.



-------------
Kevin

North Staffordshire



Posted By: spacebiscuit
Date Posted: 19 January 2024 at 6:06pm
Why are they faked - in demand and out of production?


Posted By: Xoc1
Date Posted: 19 January 2024 at 9:59pm
The 2SJ351 P channel Lateral Fet has been out of production for a long time now. and was complimentary to the 2SK2220 N channel Lateral Fet. Fakes are common as they command a good price.
Most lateral Power Fets are broadly compatible - The Exicon ECX10N20/ECX10P20. should be a good replacement. It may need a small capacitor from the gate to the voltage rail especially on the N channel to tame the Exicon as they are not as stable as the old Hitachi / Renesas parts Some examples of this capacitor soldered directly on the fet are in the amplifier rescue gods thread Posts by empro on 5th & 6th Jan 24




Posted By: EmPro
Date Posted: 20 January 2024 at 2:17pm
Even the old designs with Hitachi K135 and J50 have anti oscilating capacitors fitted on each mosfet 10pf to 22pf i've seen in most designs,in the schematic here we can see clear,it belongs to ASHLY FET2000m that uses 4 pairs of K135 and J50 per ch:
https://ibb.co/8j35hqW" rel="nofollow">



-------------
EmPro Audio


Posted By: EmPro
Date Posted: 20 January 2024 at 2:24pm
And here is the full output side,with the 4 pairs:
https://ibb.co/vv4fSYC" rel="nofollow">


-------------
EmPro Audio


Posted By: Earplug
Date Posted: 20 January 2024 at 4:10pm
Yes. I've also seen amps with little ferrite beads on the gate lead to tame the oscillation, but really not sure how effective that would be. A cap is definitely a better idea.




-------------
Earplugs Are For Wimps!


Posted By: EmPro
Date Posted: 20 January 2024 at 4:28pm
I'm agree with caps as well :) but good ones...not the ones that tends in time to become fuses LOL

-------------
EmPro Audio


Posted By: Xoc1
Date Posted: 20 January 2024 at 4:37pm
A ferrite bead on the gate lead helps, along with the capacitor, & the gate resistor should be as close to the FET as possible, not down the end of a long track! Even if the amplifier is not oscillating constantly, these strategies help, especially when replacing Old Hitachi / Renesas FET's with Exicons.


Posted By: EmPro
Date Posted: 20 January 2024 at 4:56pm
On the fet feet is the best option ever :)

-------------
EmPro Audio


Posted By: M4trix
Date Posted: 20 January 2024 at 5:17pm
You mean this ? Wink


Edit: https://forum.speakerplans.com/none-inductive-resistors_topic94087.html" rel="nofollow - https://forum.speakerplans.com/none-inductive-resistors_topic94087.html



Posted By: EmPro
Date Posted: 20 January 2024 at 5:38pm
That work's also,but i trust more the capacitors on the fet legs :)

-------------
EmPro Audio


Posted By: M4trix
Date Posted: 20 January 2024 at 10:22pm
Where is Alan ? Looks like we scared the man. Ouch


Posted By: spacebiscuit
Date Posted: 25 January 2024 at 9:16pm
I'm back - was waiting for a component tester today which I received today so I can better analysis the replacement MOSFET.

I have 2 new replacement 2SJ352, along with one of the last pair I purchased as replacements and assumed I had blown results are as follows:



All 3 that I've  tested have similar results - how do I use these results to compare to the spec on the datasheet?

https://datasheetspdf.com/pdf/201924/HitachiSemiconductor/2SJ352/1








Posted By: Earplug
Date Posted: 26 January 2024 at 7:29am
Yes, that tester won't tell you if the device is a fake and will die with a lower voltage applied.

You need something like this:-

https://es.aliexpress.com/item/1005006034179769.html?spm=a2g0o.productlist.main.21.73f0448aOV6yLn&algo_pvid=1f1de123-3cf9-4aa6-907a-e0949769a046&algo_exp_id=1f1de123-3cf9-4aa6-907a-e0949769a046-10&pdp_npi=4%40dis%21EUR%2139.69%2115.49%21%21%2142.04%2116.40%21%40211b61bb17062532853561144e4903%2112000035417422840%21sea%21ES%214785378526%21&curPageLogUid=XZD9a6Qxsal5&utparam-url=scene%3Asearch%7Cquery_from%3A" rel="nofollow - https://es.aliexpress.com/item/1005006034179769.html?spm=a2g0o.productlist.main.21.73f0448aOV6yLn&algo_pvid=1f1de123-3cf9-4aa6-907a-e0949769a046&algo_exp_id=1f1de123-3cf9-4aa6-907a-e0949769a046-10&pdp_npi=4%40dis%21EUR%2139.69%2115.49%21%21%2142.04%2116.40%21%40211b61bb17062532853561144e4903%2112000035417422840%21sea%21ES%214785378526%21&curPageLogUid=XZD9a6Qxsal5&utparam-url=scene%3Asearch%7Cquery_from%3A

Funnily enough, one friend that does a lot of SMPS repairs just bought an all singing & dancing tester from Mr. Ali as he was sick of MOSFETs failing on him. He recommended I get one. He said that it has everything on it & works a dream. I'll try get a link & post later. 





-------------
Earplugs Are For Wimps!


Posted By: Earplug
Date Posted: 26 January 2024 at 11:55am
Ok, turns out that that is the one my friend bought...   LOL

He reckons that (so far) it's working fine...

His link:-

https://es.aliexpress.com/item/1005006095568577.html?srcSns=sns_More&spreadType=socialShare&bizType=ProductDetail&social_params=60477018199&aff_fcid=128e98d2e77e4aa5b49220e85f90f25d-1706269808095-09606-_EuWlgI5&tt=MG&aff_fsk=_EuWlgI5&aff_platform=default&sk=_EuWlgI5&aff_trace_key=128e98d2e77e4aa5b49220e85f90f25d-1706269808095-09606-_EuWlgI5&shareId=60477018199&businessType=ProductDetail&platform=AE&terminal_id=e28994e1d722404fbac47df6787a938e&afSmartRedirect=y" rel="nofollow - https://es.aliexpress.com/item/1005006095568577.html?srcSns=sns_More&spreadType=socialShare&bizType=ProductDetail&social_params=60477018199&aff_fcid=128e98d2e77e4aa5b49220e85f90f25d-1706269808095-09606-_EuWlgI5&tt=MG&aff_fsk=_EuWlgI5&aff_platform=default&sk=_EuWlgI5&aff_trace_key=128e98d2e77e4aa5b49220e85f90f25d-1706269808095-09606-_EuWlgI5&shareId=60477018199&businessType=ProductDetail&platform=AE&terminal_id=e28994e1d722404fbac47df6787a938e&afSmartRedirect=y







-------------
Earplugs Are For Wimps!


Posted By: spacebiscuit
Date Posted: 29 January 2024 at 2:39pm
Currently I am trying to establish if my fuses continue to blow because of the replacment MOSFET or another damaged component.

Given that the first batch of MOSFETs which blow the fuses are giving comparable readings in my tester to the 2nd batch of MOSFETs which have not yet been installed, can I assume that they did not die with "a lower voltage applied" and the issue is another component.

Or, could a poor quality MOSFET cause the fuses to blow yet reman undamaged?

I've checked the other components on the board and I'm sure the results are comparable/similar so I'm a little bit stuck for what to do.




Posted By: Earplug
Date Posted: 29 January 2024 at 5:58pm
When in any sort of doubt and have fuses blowing, etc., you want to stick an incandescent, or halogen bulb in series with the device and save yourself some aggro.

If there is a fault, the light will come on and tell you that all is not well. I use this method all the time when switching on stuff for repair - or stuff repaired but untested. It's simple and saves fuses as well as unexpected blow-outs, fireworks, etc.   Smile




-------------
Earplugs Are For Wimps!


Posted By: spacebiscuit
Date Posted: 29 January 2024 at 7:52pm
how does this method with the bulb help - where am I adding them to my circuit. Trying to figure how they will save my fuses - won't they expose my board to the danger that the fuses are protecting?

Is this the same principle:

https://www.youtube.com/watch?v=wRFRwOnLsZI&ab_channel=UncleDoug


Posted By: Xoc1
Date Posted: 29 January 2024 at 8:40pm
A mains voltage bulb wired in line with the live feed acts as a current limiter. If there is a dead short across the amp, the bulb sees the majority of the voltage & lights up. As the bulb gets hot it resistance increases protecting the amp under test. The amplifier only sees a few volts & does not blow the fuses. You can switch off again & try to find the fault.
If the amp is not drawing excessive current, once the PSU has charged the bulb will go dim.  Fuses are really a last resort to stop the amp from catching fire - The fuses may protect you, but they don't protect the amp from failure.


Posted By: Xoc1
Date Posted: 29 January 2024 at 9:21pm
I did not find any 2SJ352 fets to test in my spares but found a 2SJ162 Hitachi which is similar. Your test results seem way off. Did you use the self test before measuring your FET's?

Note the diode from Drain to Source which is missing from your test pic!


Posted By: spacebiscuit
Date Posted: 30 January 2024 at 12:08am
Yes I did do the self-test before test - just tested again but now i am getting "unknown of damaged part" on the 2SJ352s since I put them on the board, and one of the n-MOSFET is showing as a capacitor now.

I guess the replacements were not fit for purpose and have now only made matters worse :) 

I've tested the two original 2SJ351s on the working channel (still on the board though) and they don't show the diode either but I guess it is irrelevant since they're toast anyway....

Is it time for me to trash the review for these in the hope that someone else doesn't but this fake crap?

Have no idea how I am going to source a working pair.






Posted By: Earplug
Date Posted: 30 January 2024 at 8:59am
Originally posted by spacebiscuit spacebiscuit wrote:

how does this method with the bulb help - where am I adding them to my circuit. Trying to figure how they will save my fuses - won't they expose my board to the danger that the fuses are protecting?

Is this the same principle:

https://www.youtube.com/watch?v=wRFRwOnLsZI&ab_channel=UncleDoug


Yes, that's it. Limit the current flowing through the DUT and so avoid unnecessary damage to the circuit &/or board. And avoid constantly having to replace fuses!!   LOL LOL LOL








-------------
Earplugs Are For Wimps!


Posted By: spacebiscuit
Date Posted: 30 January 2024 at 11:41am
Ok I'm going to make myself a bulb limiter - makes sense.

are the following assumptions correct:

- Would these fake crappy MOSFETs produce the behaviour I am experiencing - tripping fuses

- Will the bulb limiter protect any new MOSFET that I fit in the event that another component is damaged and also causing the circuit to trip the fuses

I've sourced what I believe to be genuine MOSFET - the seller has re-assured me that they are genuine.


Posted By: Earplug
Date Posted: 30 January 2024 at 1:00pm
Yes, my incandescent lamp limiter is an essential tool for me and also something to do with all those (now) illegal bulbs! Use at least 2/3 x 100W or equivalent.    Smile

In theory, it should protect the circuit from any damaged/fake component causing fuses blowing and further damage on the board.

Fake MOSFETs failing will definitely cause fuses to pop.






-------------
Earplugs Are For Wimps!


Posted By: spacebiscuit
Date Posted: 10 February 2024 at 12:54pm
Replacement 2SJ352 received - my tester gives this result - promising?








Posted By: APW
Date Posted: 10 February 2024 at 7:14pm
Unless these are NOS (New old stock) from a reliable source, they are almost certainly fake… there are so many fakes out there that it’s unreal and NOS stock of these is like rocking horse poop.

Unless you know the source forget about these obsolete fets… the only real viable option for new fets is using Exicon ones from profusion that M4trix posted about on the first page of this thread. 

Profusion are the main dealer for Exicon so you are guaranteed that they won’t be fake.



Posted By: spacebiscuit
Date Posted: 18 February 2024 at 8:30pm
Thanks and I do hear you regarding purchase from a reputable source. I was slightly put off by opting for the Exicon  because earlier in this thread it was suggested that I would need a mod - something I didn't really understand and thought would be beyond my skill set.

i fitted a replacement pair today, powered up via the bulb tester. After taking some voltage readings I added a source and my speakers and I was able to playback sound for about 25-30 seconds before the FETs got very hot. They are not screwed down to the heatsink as I initially wanted to make sure they weren't already toast. However even when idle without a source they tend to get quite hot quickly.

So now I'm wondering - does the heatsink work that well (on the good channel I couldn't feel any noticeable heat when touching the black from of the FET) that I simply need to screw them down or is symptomatic of another issue?

I'm not sure if this is a good sign or not, but relief to have those fuses tripping which feels like a small victory (I did power it directly to the mains without the bulb tester to confirm this).


Posted By: Xoc1
Date Posted: 18 February 2024 at 10:27pm
Definitely fix them to the heatsink with fresh thermal paste & check that there are no shorts if insulators are fitted
The Bias current should be checked which would probably involve wiring a DVM set on amps across one of the fuse holders with the fuse out.
But can't see any bias adjustment on the PCB in your pics.
Using the Exicons would not be beyond your rapidly increasing skill set. Could always help you with how to add the extra components to your PCB which might cost £1 or so!


Posted By: spacebiscuit
Date Posted: 19 February 2024 at 2:07pm
Originally posted by Xoc1 Xoc1 wrote:

check that there are no shorts if insulators are fitted

Going to screw them down, I have new paste and new pads. Can you educate me here - why would insulators cause a potential short?


Posted By: Earplug
Date Posted: 19 February 2024 at 6:25pm
Some amps are designed with rail voltages on the heatsink, so no insulators are used. That's usually done to aid cooling as devices are bolted straight onto the sink.

If the amp uses micas, then it means that the heatsinks are on the chassis, ie at earth potential. So any contact between the tab of the device & sink will end in tears.   Ouch





-------------
Earplugs Are For Wimps!


Posted By: spacebiscuit
Date Posted: 22 February 2024 at 8:49pm
Screwed the FETs down and used insulation pads and thermal paste - I have sound and believe the FETs are not getting too hot.

When I first power up (if amp has been off for a period) I sometimes get a bit of a hum that goes as I turn up the source volume and it eventually fades. I'm sometimes getting oscillation, in which case I power off and back on and it seems to disappear.

When the hum and any oscillation has gone I've let the source play for around 60-90 seconds, slowly increasing volume and the bottom end. The sound is great, nice, deep and rich, I didn't push it excessively, but enough to feel it working. Try to plan my next move....


Posted By: Earplug
Date Posted: 23 February 2024 at 10:02am
Ok. Have you stuck caps on the FET legs? 10pf-100pf, soldered between Gate & Drain. See if that helps get rid of hum & oscillation.

But good to here that sound quality is satisfying. That makes all the work worth the effort.   Thumbs Up


EDIT:-  Or it could be a problem with the main filter caps. Are they ok?







-------------
Earplugs Are For Wimps!


Posted By: spacebiscuit
Date Posted: 23 February 2024 at 10:29am
I'll give the caps a go - why do I need the caps now when previously I didn't. I'll check all other caps too a little later.

First thing to day when I fired the amp up no source connected I had quite bad distortion which steadily rose, until i powered off. Powered on again same, but this time I turned the volume a touch and I got quite a bad flapping sound. Third time just sounded like a bad earth - humming was consistent and stable - not rising. As i said - as soon as a source is added and played seems to be fine.


Posted By: Earplug
Date Posted: 23 February 2024 at 11:27am
The small caps are good to "tame" the new FETs. You never know with differences in the devices and difference between the originals & new ones.

Bad/leaky filter caps can cause all sorts of (unpredictable) problems - like distortion. Well worth changing if there is any doubt.




-------------
Earplugs Are For Wimps!


Posted By: spacebiscuit
Date Posted: 23 February 2024 at 5:37pm
Do I need the caps on both the N and P-FETS - I have two pairs.

I've I'm going to pick some up better to go for 10, 100 or something in the middle? Also do they need to be high voltage, I've seen 50v and 1kv?


Posted By: Earplug
Date Posted: 23 February 2024 at 6:15pm
Yes, best on all. Get whatever is available to you, like at your local component shop, etc. The actual value isn't that important. Maybe get a selection and try see if you can hear/note any difference. They shouldn't cost much.    LOL LOL LOL

50V is a bit low. Are there no 250V ceramic caps available? If nothing else, just get the 1kV, no problem.They are all relatively small & should fit without too much messing about. As I said, solder direct onto the legs of the FET.



-------------
Earplugs Are For Wimps!



Print Page | Close Window

Forum Software by Web Wiz Forums® version 12.06 - https://www.webwizforums.com
Copyright ©2001-2023 Web Wiz Ltd. - https://www.webwiz.net